Promotions may be included in the article

Clean 3DCG Data - What a 3DCG Beginner Has Taken Care Of In Work

Clean 3DCG Data - What a 3DCG Beginner Has Taken Care Of In Work

When I was a novice 3DCG artist, I felt guilty about my lack of skills and experience because I had just started my career as an inexperienced 3DCG artist.

In the midst of all this, if you don't have the skills or experience, at least let's do this!
I sometimes think, "I'm glad I did that," because I had decided to do it.

One of them is "making clean 3DCG data".

This article describes the benefits of clean 3DCG data and why it is important to create clean data.


Even 3DCG beginners can create beautiful data

feeling sad

For a 3DCG beginner, it is difficult to create a 3DCG that is acceptable for work.

It is not so difficult as it is impossible to make it at first.

This is especially true when you have just started working as a 3DCG artist, either as a new graduate or from a new job with no experience.

Even if you get a job at a company, you will be in training for at least three months, and even after the training is over, it is like practicing for about a year. It inevitably takes time before I can create 3DCG of satisfactory quality as a job.

That is normal, and the senior artists in the company know exactly what they are doing.


But some things can be done before you can make work-level 3DCG yet.

That is "to create clean 3DCG data".

What does clean data look like?

- Clean 3DCG Data -
  1. Unnecessary garbage data is properly deleted.
  2. Objects are named correctly.
  3. Objects do not contain unnecessary move, rotate, or scale values.
  4. Objects are in the correct position, orientation, and origin location.
  5. No overlapping vertices, no polygons, no back-facing polygons, etc.
  6. Bone axis orientation and name match specifications.
  7. The data structure is as specified in the specification.


These things have nothing to do with the look of the 3DCG.

Even if you can't make good-looking 3DCG right away, anyone can make clean data if they are careful.

You can visually check if the object names and orientations are correct. If a job has a well-written specification, you can produce clean data according to the specifications if you read them carefully.

It is difficult to check the overlap of vertices and polygons by eye, but most 3DCG software has a function to check such vertices and polygons.

If you use the check function well, even 3DCG beginners can check it by themselves.


As written in What Working Adult With No 3DCG Experience Does Became 3DCG Artist, Toha went to a technical school with no 3DCG experience and became a 3DCG artist within a year.

I became a professional 3DCG artist with the skills of a mere novice.

I knew that I was not good enough, so I worked every day with the mindset that even if I could not create good-looking 3DCG, I would at least create beautiful data.


Sponsored Links


Clean data that is unassuming but surprisingly appreciated


I know that somehow I get complimented when I make 3DCG data beautifully in this way.

Toha, who was flattered by the praise, continued to strive for "clean data" for the rest of my life.

10 years have passed, and even now that we can create 3D models of a quality that is recognized in our work, we always try to create clean data.


I have found that clean 3DCG data is valued at work.

The reason why it is valued is that There are rather many 3DCGs out that are not cleanin the world.


Only those who have actually seen the data can tell whether the 3DCG data is beautiful or not.

The people who see the data include other artists in the company, programmers who handle the data, and the client company that delivers the 3DCG created.

In other words, they are the people we work with.

It is not obvious from the outside whether the data is clean or not, and it is not something to brag about to others, so it is very low profile.

But the people I have worked with understand.

The fact that people who work with you appreciate you is the best achievement in your work.


I have said that there is a rather large amount of 3DCG data in the world that is not beautiful.

This is just a guess on Toha's part, but I believe that some 3DCG artists do not place particular importance on the cleanliness of the data.

Some people are not very good at details or at spelling their names correctly. Even if you are not a bad speller, you can still make mistakes by accident.

For a variety of reasons, it is not uncommon for 3DCG to have unclean data.

But in most cases, such 3DCG will have to be corrected later in the work.

This is because, in 3DCG work, data that is not clean is problematic.


Clean data that was important for 3DCG work


Why is data that is not clean problematic?

The main reasons are these.

- Problems with unclean 3DCG data -
  1. Difficult to understand when others look at the data
  2. When others use the data, problems arise and they cannot use it.
  3. Source of bugs in games and other programs.
  4. In the first place, an error occurs and the output cannot be output to the game.
  5. It is difficult to correct a mistake that has a wide impact later.
  6. Not very impressive to the customer to whom the data is delivered.


The work of a 3DCG artist working for a company is rarely completed on his or her own.

Some people take the work you do and then do the work for you.

Using game development work as an example, workflows like this.

- Workflow in game development -
  1. The modeler makes a 3D model.
  2. The rigger creates a rig for that 3D model.
  3. The motion artist uses that rig to create the motion.
  4. The effect artist adds effects to the motion.
  5. Output the completed data so that the programmer can play the game.



In this kind of workflow, unclean data can cause problems for those who work later in the process.

If an object or bone is misnamed, if there are strange polygons, or if there is garbage data that needs to be deleted, there will always be a problem somewhere.

If someone notices a mistake during the process, he or she can point out the error, but if the number of mistakes increases, both the person pointing out the error and the person being pointed out do not like it, and it is a hassle for both parties.

And on rare occasions, problems do not occur during the process and the 3DCG data that is not clean is used in the game. In these cases, the 3DCG data can cause unexpected bugs.


The work of a 3DCG artist is based on cooperation with the people he or she works with.

The data you create is not for your exclusive use.

So creating clean data is an important part of 3DCG work.


Clean data that benefits you as well


Clean 3DCG data is good for those around you who work with you.

And making your data clean benefits not only those around you but also yourself.

- Advantages of creating clean 3DCG data -
  1. Clean data is appreciated by those around you.
  2. Clean data leads to trust in yourself.
  3. Saves time in correcting incorrect data.


Again, clean data is appreciated on its own.

It is much easier than making good 3DCG, but it is also an area that is often neglected because it is too much of a hassle or because people are too busy to spend time on it.

So if you produce clean data every time you work, you will gradually gain the trust of those around you.

More often than not, people think, "This person always produces clean data, so I can rest assured.


Well, even so, we are human beings and we can make mistakes.

No matter how careful we are, we all make mistakes. When this happens, the only thing you can do is to apologize, correct the data, and be careful not to make the same mistake again.

However, if you are careful, you can certainly reduce the number of mistakes. People who are careful to reduce the number of mistakes are, in the eyes of others, "people who make fewer mistakes.

The fewer data errors there are, the less time and effort is required to correct the data.



By the way, "clean data" here refers to "finalized 3DCG data". There is no problem with data in the process of creation, as long as you know what you are doing.

However, we recommend that you at least clean up the data during the creation process.


The minimum clean data is "enough data that I can look at it a week later and understand it".

For example, if all the object names are cube01, cube02, cube03, etc., when you open the data a week later, you will not know which object name was which.

You should at least give the objects in your scene names that you can recognize.

It may seem like a hassle to name objects every time, but if you don't name them, it will be even more of a hassle later. Please make it a habit to name your objects.


Sponsored Links


The key to creating beautiful 3DCG data is to be aware & not be in a hurry


Now, you say that clean data can be produced if we are careful, but how exactly do we take care of it?

The first and most primitive way is to make a checklist.

Before submitting the completed data, you go through the checklist to make sure the data is clean. The checklist should include the following items, which are summarized in the summary section of this article.

Even without a checklist, simple data errors can be reduced by "checking the completed data carefully one last time yourself".


First and foremost, be conscious of trying to create clean data is important.

If you are not comfortable with detailed checking, you can use a script to check automatically. (I don't know much about scripts, so I'll spare you the details.)


And quite importantly, don't rush to create data.

In Toha's experience, there is an 80% probability that something is wrong with the data created in a hurry.

When the deadline for submitting data is just around the corner, or when you want to finish work early and go home, the data created under these circumstances is usually incorrect, and you will have to correct it later.

It is easy to make mistakes when you are in a hurry.

When creating data to be implemented in a game or submitted to a client, we recommend that you first work calmly, unhurriedly, and carefully.


Summary: There is nothing to lose by creating clean 3DCG data


In this article, I wrote about "creating beautiful 3DCG data," something I have been glad to have done since Toha was a novice 3DCG artist.

Clean data is data that looks like this.

- Clean 3DCG Data -
  1. Unnecessary garbage data is properly deleted.
  2. Objects are named correctly.
  3. Objects do not contain unnecessary move, rotate, or scale values.
  4. Objects are in the correct position, orientation, and origin location.
  5. No overlapping vertices, no polygons, no back-facing polygons, etc.
  6. Bone axis orientation and name match specifications.
  7. The data structure is as specified in the specification.


For Maya data, this includes making sure the history is properly erased.

Toha's conclusion from his experience is that creating beautiful 3DCG data is "properly valued as the work of a 3DCG artist".

No one is bothered by clean data. Rather, everyone is happy and it benefits me.

If you want to work in 3DCG and have never thought about the cleanliness of data, please be conscious of creating clean 3DCG data from now on.


It will be good!


Actually, there is one more thing that I am glad I continued to do after becoming a 3DCG artist.

It is to "record the details of the 3DCG production work every day.

For this, I wrote Results Of Daily Recording In 3DCG Work - Work Records Are Full Of Useful Information! if you like!

List of comments

  1. しまりす says:

    Making it beautiful is important.
    Daily noteI have been slacking lately, but I decided to keep a record again.

    • Toha トハ says:

      Chipmunks, thank you for your comments~!
      It will be quite helpful for me later on if I make the data clean.
      I think it would be useful to keep a note of it every day.


Please feel free to comment~!

Copied title and URL